Super Street Fighter IV Arcade Edition: Trial Fail

So I got SSFIV:AE and I’m breaking in my Saturn USB pad by playing the Trial mode.

The Trial mode consists of a series of stages. In each of those, the player has to execute one or multiple moves. It starts with the simple execution of a special move and quickly moves on to combos of increasing difficulty. While it sounds like a reasonable learning curve, the concrete implementation doesn’t turn out to be as helpful as it can be.

But let’s start with the minor issues.

For all you busy people: I’ve outlined an alternative concept for game-assisted training at the end.

Tatsumaki is not a kind of sushi

We’ll start with the really obvious: the required move(s) in each level is not shown as commands, but by their names. For a mode made for structured growth (as stated by Capcom), it is safe to assume that the names are rather meaningless to the novice player — even if he has heard of them before. The many Japanese names bring additional difficulty to the identification, differentiation and memorization of the terms with the associated moves. You might know the “Shoryuken motion” (which cannot be said for some of my game developer fellows), but terms like “Tatsumaki Senpukyaku” or even the readable “Corkscrew Cross” simply has zero use when learning new characters.

Capcom’s Seth Killian making this point for me: “Collarbone Breaker… What is that?” (0:42)

The game offers to show the input in a modal menu hidden behind a button press, but it is not an adequate solution, as the screen-covering menu removes the player from the context where the shown data is actually needed.


Click to see my mockup in a higher resolution.

Actually the original SF4 already did it better. It featured toggling between the name and the command display without going into a modal screen like SSF4. Capcom, take note: iterative design is supposed to improve things. Personally I think showing the command next to the name would be even better. Command display can well be optional since names are easier to handle in mind and in discussion, but it is only useful when you know what they refer to in the first place.

A little less conversation, a little more action please

Each stage is automatically closed right after one successful execution, so you cannot improve that execution by repeating the newly learned move until you really get it. Especially for combo sequences with more difficult timings (e.g. links, FADC to Ultra), the immediate, forced level conclusion is really counterproductive. If you use the retry option when concluding a stage, the net playtime is shorter than the interruption, which is just absurd.

9 seconds are between “success” and the next “go!” (if you proceed immediately, selecting “retry” will take a tiny bit longer). A Shoryuken is done in less than 1 second.

In the original Street Fighter IV, Trial objectives are grouped. So you are tasked with, say, 5 different objectives in one “round” before the game interrupts you with a menu, while the Super/Arcade edition interrupts you after every objective. Again, Capcom, this is not how you iterate.

One could argue that the Training stage can be used for repetition practice. While that is certainly true, having to “resort” to the free Training mode just proves how Capcom missed the mark. As is, I see only one use for the Trial mode: the step-by-step confirmation of combos.

I’m not offering a solution at this point, but please read on — this issue is addressed at the end in a more comprehensive way.

Let the monkey see so it can do

One of the early challenges are Links, which require precise timing so the subsequent attack connects before the opponent recovers from the first hit. (For those who don’t know Links, here’s a quick rundown of the different combo types.)

Contrary to the “regular” combos — which are relatively easy thanks to buffered inputs and move canceling — entering commands as quickly as possible will not help with Links. You have to get the “rhythm” of Links and press buttons at very specific moments in time. The problem is that the novice learner isn’t told how that rhythm is. There is neither a demonstration, nor a visual reference for the timing.

Another version of the same problem involves command counters. Command counters put the player character into a temporary “trap”-state which respond to incoming attacks with an automatic retaliation. So for the command counter to succeed, the dummy opponent repeatedly tries to hit you, and you have to trigger the “trap” move just before that.

Here the player isn’t informed when to do the move either. Theoretically the opponent attacks in a constant rhythm, but when you end up blocking or receiving the incoming attack — as you haven’t mastered the move yet — you also mess up the rhythm that the dummy opponent has established, making his next attempt as unpredictable as the very first one.

In both cases, your early successes result from a good deal of experiment and luck. However, the Trial mode makes it hard to convert the “accidental” successes into sustained proficiency, since each stage is immediately concluded after one successful execution. I will discuss this topic later in more detail.


Even Tekken 3 did something better.

For links, a timeline or a demonstration would help tremendously — already Tekken 3 from 1997 (maybe even its prequels) used peep-sounds to demonstrate the input timingin its combo training mode.

For command counters, a simple countdown or a visual “announcement” would have solved the problem. A working example can be found in Bayonetta, where enemy attacks are telegraphed with an audio-visual cue.

Curriculum fail

I was surprised by when the programme asks the player to do Links — right after combos by canceling. Due to the relatively forgiving timing, combos by canceling are easily discovered by quick inputs during play. As we all know, self-discovered knowledge stick the best, so buffered inputs and canceling will arguably be the “default” behavior to the player.

In addition to this mindset, scheduling combos by canceling right before the Links just builds up wrong expectations. When Link combos suddenly asks for a totally different timing, the learning player is left wondering why what has always worked stopped working at once.

Maybe Capcom’s designers just intended to keep the player on his toes — it is a “Trial” in the “Challenge” mode after all. However, the player has no means (inside the game) to learn why the usual timing no longer works. Plus, not building upon the previous step really reduces the objectives to one-offs that are useless in the game proper.

Speaking of proper fights, the strict timing raises the more fundamental question: how am I supposed to really use those combos for real, if even one execution in controlled environment is giving me such a hard time?

The answer is, of course, practice, up to the point where execution no longer requires much conscious thought, like shifting the stick when driving. Too bad the Trial mode interruptions work against mechanical mastery by repetition.

Beyond the mechanical, the really interesting challenge is knowing when to use a given move. Of course this skill is best honed in real fights, but I do have an idea for training both the confident execution and situational reaction at the same time.

*drum roll*

The Dojo Mode

The idea is to ask for each objective in live matches against the CPU, as if in the arcade mode. However, only the current objective will count as damage. The player is granted infinite health for unlimited training, while the opponent is only susceptible to the required attack, and will sustain X times of it.

Why is this good? For starters, the objective can be practiced repeatedly in a more realistic context. Plus, the player can and must learn when to execute the required attack multiple times.

More significantly, a more nuanced learning can be achieved by tying the opponent AI level to the times of successful execution. For example, the opponent starts out standing still. After some hits, he will move around, so the player can better learn the spacing dynamically. Then, with more hits, the opponent becomes more active and intelligent, requiring more effort and proficiency from the player in using the required move.

Click the image to see my mockup in a higher resolution.

So, at the beginning the player can focus on the input. When he has mastered it undisturbed, he can start worrying about doing it at the right time.

For an even better learning effect, assessment and adaptive difficulty are included in this concept “for free”: if the player fails to consistently land the required attack at a given AI level, the AI simply goes back one level, not unlike sport leagues.

In other words, each objective, be it a simple move or a combo sequence, becomes a experience bar that the player can level up individually. The various “talents”, if you will, would also clearly show one’s strengths and weaknesses, further supporting the self-assessment on a broader scale.

Click the image to see my mockup in a higher resolution.
Please excuse the poor art due to the lack of proper assets to play with.

With this concept, we not only solve the problems of repetition and real-world application; the adaptive training style also adds to each objective a learning curve that’s fun and meaningful to progress in, because it allows gradual growth and reflects the player’s proficiency in a much more realistic way. Extending this concept beyond special moves and combos would have the added benefit of shifting emphasis to situational awareness — which is what really matters in real fights — as even basic moves can trained in depth and to great effect.

Super Street Fighter IV Arcade Edition

Yu-Chung Chen

Yu-Chung Chen is a designer working primarily on video games. He studied at Köln International School of Design and has contributed to a number of published games. Currently he works as a freelance UI designer at Keen Games.

10 responses to “Super Street Fighter IV Arcade Edition: Trial Fail”

  1. Rufus

    It would be a good thing to have the trial mode extensible, allow people to upload trials to an internet database, and, ideally, support things like hit-confirms and option selects.

    1. Yu-Chung Chen

      I see you are more versed in the intricacies of the system — I haven’t even thought about advanced techniques/strategies such as those you mentioned. I would definitely want to practice those too.

      Social interactions can be a powerful motivator, so that sounds like a good idea too. A platform for exchanging training scenarios would need some careful design to ensure the content is manageable in scope and sufficient in quality. Fodder for another post ;)

  2. Krystian Majewski

    Some really sharp observations, as always. I’m actually surprised that the Trial mode is still so unusable. When it first came out if SF4, it looked like a promising feature that had the potential to be expanded into a really helpful system. It’s sad that Capcom missed that chance so many times.

    I was wondering, do you base your assumptions about how people learn on any specific pedagogic theory?

    1. Yu-Chung Chen

      No, and actually I didn’t think about it — just saw things I don’t like and tried to improve, the Dojo concept was even developed only during the writing process. So, I haven’t really examined things in a fundamental way.

      Are you seeing any assumptions that should not be taken granted?

      1. Krystian Majewski

        Oh! Not at all. On the contrary. It kinda sounded like your thesis research helped you recognizing this.

        1. Yu-Chung Chen

          I guess you’re right! I didn’t consciously think about that paper but some aspects must have stuck in my mind. Maybe my Wing Chun learning also plays a role ;-)

  3. Billy Bissette

    I don’t know if it is true, but I remember hearing at the time that Namco had patented some aspect of Tekken’s training mode. This was given as a reason for why other fighters didn’t directly copy Tekken’s combo training system.

    Considering some of the completely ludicrous videogame and computer things that successfully received patents, it wouldn’t be impossible.

  4. Andy Mangaka

    Try the trainning “trial” mode on Virtua Fighter 4 Evolution

    That’s how a trainning mode must be…

  5. bionice


  6. Kernel Panic - Sciences & cultures numériques

    [...] est le point commun entre The Elder Scrolls V : Skyrim, Mass Effect, Street Fighter IV, Worms ou encoreDiablo 3 ? Ce sont tous de grands jeux acclamés par les critiques et le public [...]


Game Design Reviews is a Blog used by a group of game designers from Germany to publish and discuss their thoughts on various games. The blog consists entirely of reviews of games. Each review focuses on the important game design ideas and concepts of that particular game. We also run a second, more informal Blog called Game Design Scrapbook.


follow Krystian on Twitter
follow Yu-Chung on Twitter
follow Daniel on Twitter